INDEX

Explanations

occurrences related to a specific point in time, which is "the past"

oai_token-act-pair · gpt-3.5-turbo

references to past events or time periods

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wagon

-0.75

 Franch

-0.73

atism

-0.71

starting

-0.70

hat

-0.70

rolling

-0.67

BuyableInstoreAndOnline

-0.67

utters

-0.66

hyde

-0.65

ithmetic

-0.64

POSITIVE LOGITS

ebin

1.48

 tense

1.27

decade

1.13

few

0.93

orate

0.90

 week

0.89

 month

0.89

 fortnight

0.86

 century

0.86

 year

0.85

Activations Density 0.028%

occurrences related to a specific point in time, which is "the past"

references to past events or time periods

No Comments

No Known Activations